is depthimage_to_laserscan limited to forward facing cameras?

asked 2018-01-28 23:30:16 -0600

dcurtis gravatar image

The documentation for pointcloud_to_laserscan says:

"If you're trying to create a virtual laserscan from your RGBD device, and your sensor is forward-facing, you'll find depthimage_to_laserscan will be much more straightforward and efficient..."

The forward-facing requirement for depthimage_to_laserscan is non-intuitive. Won't the TF take care of that?

edit retag flag offensive close merge delete

Comments

Don't think it is required, but usually it is a good idea for the same reason you look forward most of the time when driving a car, riding or walking. Also please note that not all nodes (amcl?) are able to handle multiple sensors, e.g. one front one rear.

Humpelstilzchen gravatar image Humpelstilzchen  ( 2018-01-29 02:08:09 -0600 )edit

@dcurtis:

you'll find depthimage_to_laserscan will be much more straightforward and efficient

not a native speaker, but to me this sentence conveys the fact that depthimage_to_laserscan is a good candidate node in that case, not that it requires forward-facing cameras.

gvdhoorn gravatar image gvdhoorn  ( 2018-01-29 02:16:30 -0600 )edit

I think the key is in the sentence that follow in the doc: "(...) However, if your sensor is angled, or you have some other esoteric use case, you may find this node (pointcloud_to_laserscan) to be very helpful!"

Humpelstilzchen gravatar image Humpelstilzchen  ( 2018-01-31 00:48:07 -0600 )edit